Development and evaluation of alternative models for forecasting church tax revenue

Building on the results of a research project led by Professor Dr. Gerhard Wagenhals, University of Hohenheim, in which time-series-analytical ARIMA models were used to forecast church tax revenue, this project aimed to use structural and, in particular, economic indicators to supplement the forecast of church tax revenue for the regional church in Württemberg.

Various alternative methods were systematically evaluated on the basis of ex-post forecasts. It was shown that the additional consideration of systematic and, in particular, economic factors in suitably specified econometric time series models allows a further improvement in the short-term forecast quality. In particular, the economic movements that are clearly noticeable in church tax revenue are incorporated into the forecasts faster and earlier than in models that are limited to information from past church tax revenue.